Method #1. Tree-based cllassification

Step 1 : Collecting the data

import pandas as pd
input_file = "/Users/hit028/Desktop/530 Lab1/credit.csv"
credit = pd.read_csv(input_file)

Step 2 : Exploring the data

credit.describe()

	Creditability	Account Balance	Duration of Credit (month)	Payment Status of Previous Credit	Purpose	Credit Amount	Value Savings/Stocks	Length of current employment	Instalment per cent	Sex & Marital Status	...	Duration in Current address	Most valuable available asset	Age (years)	Concurrent Credits	Type of apartment	No of Credits at this Bank	Occupation	No of dependents	Telephone	Foreign Worker
count	1000.000000	1000.000000	1000.000000	1000.00000	1000.000000	1000.00000	1000.000000	1000.000000	1000.000000	1000.00000	...	1000.000000	1000.000000	1000.00000	1000.000000	1000.000000	1000.000000	1000.000000	1000.000000	1000.000000	1000.000000
mean	0.700000	2.577000	20.903000	2.54500	2.828000	3271.24800	2.105000	3.384000	2.973000	2.68200	...	2.845000	2.358000	35.54200	2.675000	1.928000	1.407000	2.904000	1.155000	1.404000	1.037000
std	0.458487	1.257638	12.058814	1.08312	2.744439	2822.75176	1.580023	1.208306	1.118715	0.70808	...	1.103718	1.050209	11.35267	0.705601	0.530186	0.577654	0.653614	0.362086	0.490943	0.188856
min	0.000000	1.000000	4.000000	0.00000	0.000000	250.00000	1.000000	1.000000	1.000000	1.00000	...	1.000000	1.000000	19.00000	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
25%	0.000000	1.000000	12.000000	2.00000	1.000000	1365.50000	1.000000	3.000000	2.000000	2.00000	...	2.000000	1.000000	27.00000	3.000000	2.000000	1.000000	3.000000	1.000000	1.000000	1.000000
50%	1.000000	2.000000	18.000000	2.00000	2.000000	2319.50000	1.000000	3.000000	3.000000	3.00000	...	3.000000	2.000000	33.00000	3.000000	2.000000	1.000000	3.000000	1.000000	1.000000	1.000000
75%	1.000000	4.000000	24.000000	4.00000	3.000000	3972.25000	3.000000	5.000000	4.000000	3.00000	...	4.000000	3.000000	42.00000	3.000000	2.000000	2.000000	3.000000	1.000000	2.000000	1.000000
max	1.000000	4.000000	72.000000	4.00000	10.000000	18424.00000	5.000000	5.000000	4.000000	4.00000	...	4.000000	4.000000	75.00000	3.000000	3.000000	4.000000	4.000000	2.000000	2.000000	2.000000

8 rows × 21 columns

credit.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1000 entries, 0 to 999
Data columns (total 21 columns):
Creditability                        1000 non-null int64
Account Balance                      1000 non-null int64
Duration of Credit (month)           1000 non-null int64
Payment Status of Previous Credit    1000 non-null int64
Purpose                              1000 non-null int64
Credit Amount                        1000 non-null int64
Value Savings/Stocks                 1000 non-null int64
Length of current employment         1000 non-null int64
Instalment per cent                  1000 non-null int64
Sex & Marital Status                 1000 non-null int64
Guarantors                           1000 non-null int64
Duration in Current address          1000 non-null int64
Most valuable available asset        1000 non-null int64
Age (years)                          1000 non-null int64
Concurrent Credits                   1000 non-null int64
Type of apartment                    1000 non-null int64
No of Credits at this Bank           1000 non-null int64
Occupation                           1000 non-null int64
No of dependents                     1000 non-null int64
Telephone                            1000 non-null int64
Foreign Worker                       1000 non-null int64
dtypes: int64(21)
memory usage: 164.2 KB

target = credit['Creditability']
target.value_counts()

1    700
0    300
Name: Creditability, dtype: int64

import random
random.seed(12345)
indx = random.sample(range(0, 1000), 1000)
credit_rand = credit.iloc[indx]
target_rand = target.iloc[indx]

credit_rand.head(5)

	Creditability	Account Balance	Duration of Credit (month)	Payment Status of Previous Credit	Purpose	Credit Amount	Value Savings/Stocks	Length of current employment	Instalment per cent	Sex & Marital Status	...	Duration in Current address	Most valuable available asset	Age (years)	Concurrent Credits	Type of apartment	No of Credits at this Bank	Occupation	No of dependents	Telephone	Foreign Worker
426	1	2	39	3	6	11760	2	4	2	3	...	3	4	32	3	1	1	3	1	2	1
750	1	2	36	3	0	2862	2	5	4	3	...	3	4	30	3	3	1	3	1	1	1
10	1	1	11	4	0	3905	1	3	2	3	...	2	1	36	3	1	2	3	2	1	1
839	0	4	12	2	0	1386	3	3	2	2	...	2	2	26	3	2	1	3	1	1	1
845	0	2	27	4	3	2520	3	3	4	3	...	2	2	23	3	2	2	2	1	1	1

5 rows × 21 columns

credit_rand.describe()

	Creditability	Account Balance	Duration of Credit (month)	Payment Status of Previous Credit	Purpose	Credit Amount	Value Savings/Stocks	Length of current employment	Instalment per cent	Sex & Marital Status	...	Duration in Current address	Most valuable available asset	Age (years)	Concurrent Credits	Type of apartment	No of Credits at this Bank	Occupation	No of dependents	Telephone	Foreign Worker
count	1000.000000	1000.000000	1000.000000	1000.00000	1000.000000	1000.00000	1000.000000	1000.000000	1000.000000	1000.00000	...	1000.000000	1000.000000	1000.00000	1000.000000	1000.000000	1000.000000	1000.000000	1000.000000	1000.000000	1000.000000
mean	0.700000	2.577000	20.903000	2.54500	2.828000	3271.24800	2.105000	3.384000	2.973000	2.68200	...	2.845000	2.358000	35.54200	2.675000	1.928000	1.407000	2.904000	1.155000	1.404000	1.037000
std	0.458487	1.257638	12.058814	1.08312	2.744439	2822.75176	1.580023	1.208306	1.118715	0.70808	...	1.103718	1.050209	11.35267	0.705601	0.530186	0.577654	0.653614	0.362086	0.490943	0.188856
min	0.000000	1.000000	4.000000	0.00000	0.000000	250.00000	1.000000	1.000000	1.000000	1.00000	...	1.000000	1.000000	19.00000	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000	1.000000
25%	0.000000	1.000000	12.000000	2.00000	1.000000	1365.50000	1.000000	3.000000	2.000000	2.00000	...	2.000000	1.000000	27.00000	3.000000	2.000000	1.000000	3.000000	1.000000	1.000000	1.000000
50%	1.000000	2.000000	18.000000	2.00000	2.000000	2319.50000	1.000000	3.000000	3.000000	3.00000	...	3.000000	2.000000	33.00000	3.000000	2.000000	1.000000	3.000000	1.000000	1.000000	1.000000
75%	1.000000	4.000000	24.000000	4.00000	3.000000	3972.25000	3.000000	5.000000	4.000000	3.00000	...	4.000000	3.000000	42.00000	3.000000	2.000000	2.000000	3.000000	1.000000	2.000000	1.000000
max	1.000000	4.000000	72.000000	4.00000	10.000000	18424.00000	5.000000	5.000000	4.000000	4.00000	...	4.000000	4.000000	75.00000	3.000000	3.000000	4.000000	4.000000	2.000000	2.000000	2.000000

8 rows × 21 columns

credit_train = credit_rand.iloc[0:70]
credit_test = credit_rand.iloc[700:1000]
target_train = target_rand.iloc[0:700]
target_test = target_rand.iloc[700:1000]

target_train.value_counts()/700

1    0.691429
0    0.308571
Name: Creditability, dtype: float64

target_test.value_counts()/300

1    0.72
0    0.28
Name: Creditability, dtype: float64

from sklearn.model_selection import train_test_split

y = target
X = credit.drop(['Creditability'], axis=1)

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.30, random_state = 52)

Step 3 : Training a model on the data

from sklearn import tree
from sklearn.tree import DecisionTreeClassifier

model = tree.DecisionTreeClassifier()
model = model.fit(X_train, y_train)

from IPython.display import SVG

pip install graphviz

Requirement already satisfied: graphviz in /Users/hit028/opt/anaconda3/lib/python3.7/site-packages (0.13.2)
Note: you may need to restart the kernel to use updated packages.

from graphviz import Source
from IPython.display import display

graph = Source(tree.export_graphviz(model, out_file=None, feature_names=X.columns, class_names=['default', 'no default'], filled = True))

conda install graphviz

Collecting package metadata (current_repodata.json): done
Solving environment: done

# All requested packages already installed.


Note: you may need to restart the kernel to use updated packages.

display(SVG(graph.pipe(format = 'svg')))

Step 4 : Evaluating model performance

from sklearn.metrics import confusion_matrix
from sklearn.metrics import accuracy_score

y_predict = model.predict(X_test)

confusion_matrix(y_test, y_predict)
accuracy_score(y_test, y_predict)*100

68.33333333333333

Q1 : If you see the accuracy 0f 100%, what does it mean? Does this mean that we design a perfect model? This is some thing that needs more discussion. Write a few sentences about accuracy of 100%.

Answer : It doesn't mean that the model is perfect when the accuracy score is 100% because it's probably too good to be true and might be overfitted. In the real world it 100% accuracy doesn't sound true...

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.30, random_state=23458)

Method 2 : Random Forest

from sklearn.ensemble import RandomForestClassifier
clf = RandomForestClassifier() 
clf.fit(X_train, y_train)

/Users/hit028/opt/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
  "10 in version 0.20 to 100 in 0.22.", FutureWarning)

RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',
                       max_depth=None, max_features='auto', max_leaf_nodes=None,
                       min_impurity_decrease=0.0, min_impurity_split=None,
                       min_samples_leaf=1, min_samples_split=2,
                       min_weight_fraction_leaf=0.0, n_estimators=10,
                       n_jobs=None, oob_score=False, random_state=None,
                       verbose=0, warm_start=False)

y_predict = clf.predict(X_test) 
print(confusion_matrix(y_test, y_predict))

[[ 52  30]
 [ 39 179]]

print(accuracy_score(y_test, y_predict)*100)

77.0

Q2 : What are the three most important features in this model?

Answer : From below, they are "purpose", "creditability", and "account balance". However, when I run this command many times, it sometimes show that the 3rd importance feature as "age".

import pandas as pd

credit.head()

	Creditability	Account Balance	Duration of Credit (month)	Payment Status of Previous Credit	Purpose	Credit Amount	Value Savings/Stocks	Length of current employment	Instalment per cent	Sex & Marital Status	...	Duration in Current address	Most valuable available asset	Age (years)	Concurrent Credits	Type of apartment	No of Credits at this Bank	Occupation	No of dependents	Telephone	Foreign Worker
0	1	1	18	4	2	1049	1	2	4	2	...	4	2	21	3	1	1	3	1	1	1
1	1	1	9	4	0	2799	1	3	2	3	...	2	1	36	3	1	2	3	2	1	1
2	1	2	12	2	9	841	2	4	2	2	...	4	1	23	3	1	1	2	1	1	1
3	1	1	12	4	0	2122	1	3	3	3	...	2	1	39	3	1	2	2	2	1	2
4	1	1	12	4	0	2171	1	3	4	3	...	4	2	38	1	2	2	2	1	1	2

5 rows × 21 columns

cols = ["creditability", "account_balance", "duration", "payment_status", "purpose", "credit_amount", "value", "employment", "instalment", "sex", "current_address", "available_asset", "age", "credits", "apartment", "this_bank", "occupation", "dependents", "telephone", "foreign_worker"]

feature_importances = pd.DataFrame(clf.feature_importances_, index = cols[0:20], columns=['importance']).sort_values('importance', ascending=False) 

feature_importances

	importance
purpose	0.141127
age	0.113862
creditability	0.107613
account_balance	0.107119
payment_status	0.062547
duration	0.053750
value	0.049357
available_asset	0.046402
credit_amount	0.043455
employment	0.042214
current_address	0.037916
occupation	0.034961
credits	0.029641
this_bank	0.028420
apartment	0.028033
instalment	0.027109
telephone	0.020021
dependents	0.013149
sex	0.011470
foreign_worker	0.001833

Method 3 : Adding regression to trees

Step 1 : Collecting the data

input_file2 = "/Users/hit028/Desktop/530 Lab1/whitewines.csv"

wine = pd.read_csv(input_file2)
wine.head(5)

	fixed acidity	volatile acidity	citric acid	residual sugar	chlorides	free sulfur dioxide	total sulfur dioxide	density	pH	sulphates	alcohol	quality
0	6.7	0.62	0.24	1.10	0.039	6.0	62.0	0.99340	3.41	0.32	10.400000	5
1	5.7	0.22	0.20	16.00	0.044	41.0	113.0	0.99862	3.22	0.46	8.900000	6
2	5.9	0.19	0.26	7.40	0.034	33.0	123.0	0.99500	3.49	0.42	10.100000	6
3	5.3	0.47	0.10	1.30	0.036	11.0	74.0	0.99082	3.48	0.54	11.200000	4
4	6.4	0.29	0.21	9.65	0.041	36.0	119.0	0.99334	2.99	0.34	10.933333	6

wine.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 4898 entries, 0 to 4897
Data columns (total 12 columns):
fixed acidity           4898 non-null float64
volatile acidity        4898 non-null float64
citric acid             4898 non-null float64
residual sugar          4898 non-null float64
chlorides               4898 non-null float64
free sulfur dioxide     4898 non-null float64
total sulfur dioxide    4898 non-null float64
density                 4898 non-null float64
pH                      4898 non-null float64
sulphates               4898 non-null float64
alcohol                 4898 non-null float64
quality                 4898 non-null int64
dtypes: float64(11), int64(1)
memory usage: 459.3 KB

import matplotlib.pyplot as plt

n, bins, patches = plt.hist(x=wine['quality'], bins='auto', color='b', )
plt.xlabel('Quallity Value')
plt.ylabel('Frequency')
plt.show()

Step 2 : Exploring and preparing the data

target2 = wine['quality']
target2.value_counts()

6    2198
5    1457
7     880
8     175
4     163
3      20
9       5
Name: quality, dtype: int64

from sklearn.model_selection import train_test_split

y2 = target2 
X2 = wine.drop(['quality'], axis=1)

X2_train, X2_test, y2_train, y2_test = train_test_split(X2, y2, test_size=0.25, random_state=1234)

Step 3 : Splitting Data and train the model

from sklearn import tree
from sklearn.tree import DecisionTreeClassifier

model2 = tree.DecisionTreeClassifier()
model2 = model2.fit(X2_train, y2_train)

from sklearn.tree import export_graphviz

export_graphviz(model2, out_file = 'tree.dot', feature_names = X2.columns)

from graphviz import Source 

dot_data2 = tree.export_graphviz(model2, out_file = None, feature_names = X2.columns, filled = True, rounded = True, special_characters = True)

import graphviz 
dot_data2= tree.export_graphviz(model2, out_file=None) 
graph2 = graphviz.Source(dot_data2)

graph2 = graphviz.Source(dot_data2)

graph2

Step 4 : Evaluating Model Performance

import numpy as np

np.corrcoef(y_test,y_predict)

array([[1.        , 0.44135563],
       [0.44135563, 1.        ]])

Q3 : What is your interpretation about this amount of RMSE?

Answer : Since it's the error value, we want it to be low as possible, therefore this RMSE value indicates this is not quite ideal...

Method #4. News Popularity

Step 1 : Collecting the data

input_file3 = "/Users/hit028/Desktop/530 Lab1/OnlineNewsPopularityP.csv"

import pandas as pd
news = pd.read_csv(input_file3)
news.head(5)

	url	timedelta	n_tokens_title	n_tokens_content	n_unique_tokens	n_non_stop_words	n_non_stop_unique_tokens	num_hrefs	num_self_hrefs	num_imgs	...	min_positive_polarity	max_positive_polarity	avg_negative_polarity	min_negative_polarity	max_negative_polarity	title_subjectivity	title_sentiment_polarity	abs_title_subjectivity	abs_title_sentiment_polarity	shares
0	http://mashable.com/2013/01/07/amazon-instant-...	731	12	219	0.663594	1.0	0.815385	4	2	1	...	0.100000	0.7	-0.350000	-0.600	-0.200000	0.500000	-0.187500	0.000000	0.187500	593
1	http://mashable.com/2013/01/07/ap-samsung-spon...	731	9	255	0.604743	1.0	0.791946	3	1	1	...	0.033333	0.7	-0.118750	-0.125	-0.100000	0.000000	0.000000	0.500000	0.000000	711
2	http://mashable.com/2013/01/07/apple-40-billio...	731	9	211	0.575130	1.0	0.663866	3	1	1	...	0.100000	1.0	-0.466667	-0.800	-0.133333	0.000000	0.000000	0.500000	0.000000	1500
3	http://mashable.com/2013/01/07/astronaut-notre...	731	9	531	0.503788	1.0	0.665635	9	0	1	...	0.136364	0.8	-0.369697	-0.600	-0.166667	0.000000	0.000000	0.500000	0.000000	1200
4	http://mashable.com/2013/01/07/att-u-verse-apps/	731	13	1072	0.415646	1.0	0.540890	19	19	20	...	0.033333	1.0	-0.220192	-0.500	-0.050000	0.454545	0.136364	0.045455	0.136364	505

5 rows × 61 columns

Step 2 : Pre-processing

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns

target3 = news['shares']

target3.value_counts()

1100      2308
1200      2018
1300      1740
1400      1592
1500      1323
          ... 
197600       1
310800       1
92           1
57800        1
82200        1
Name: shares, Length: 1454, dtype: int64

popular = news.shares >= 1400
unpopular = news.shares < 1400 
news.loc[popular,'shares'] = 1 
news.loc[unpopular,'shares'] = 0

import random
random.seed(12345)
indx3 = random.sample(range(0, 1000), 1000) 
news_rand = news.iloc[indx3] 
target3_rand = target3.iloc[indx3]

news_train = news_rand.iloc[0:700] 
news_test = news_rand.iloc[700:1000]
target3_train = target3_rand.iloc[0:700] 
target3_test = target3_rand.iloc[700:1000]

news.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 39644 entries, 0 to 39643
Data columns (total 61 columns):
url                              39644 non-null object
timedelta                        39644 non-null int64
n_tokens_title                   39644 non-null int64
n_tokens_content                 39644 non-null int64
n_unique_tokens                  39644 non-null float64
n_non_stop_words                 39644 non-null float64
n_non_stop_unique_tokens         39644 non-null float64
num_hrefs                        39644 non-null int64
num_self_hrefs                   39644 non-null int64
num_imgs                         39644 non-null int64
num_videos                       39644 non-null int64
average_token_length             39644 non-null float64
num_keywords                     39644 non-null int64
data_channel_is_lifestyle        39644 non-null int64
data_channel_is_entertainment    39644 non-null int64
data_channel_is_bus              39644 non-null int64
data_channel_is_socmed           39644 non-null int64
data_channel_is_tech             39644 non-null int64
data_channel_is_world            39644 non-null int64
kw_min_min                       39644 non-null int64
kw_max_min                       39644 non-null float64
kw_avg_min                       39644 non-null float64
kw_min_max                       39644 non-null int64
kw_max_max                       39644 non-null int64
kw_avg_max                       39644 non-null float64
kw_min_avg                       39644 non-null float64
kw_max_avg                       39644 non-null float64
kw_avg_avg                       39644 non-null float64
self_reference_min_shares        39644 non-null float64
self_reference_max_shares        39644 non-null float64
self_reference_avg_sharess       39644 non-null float64
weekday_is_monday                39644 non-null int64
weekday_is_tuesday               39644 non-null int64
weekday_is_wednesday             39644 non-null int64
weekday_is_thursday              39644 non-null int64
weekday_is_friday                39644 non-null int64
weekday_is_saturday              39644 non-null int64
weekday_is_sunday                39644 non-null int64
is_weekend                       39644 non-null int64
LDA_00                           39644 non-null float64
LDA_01                           39644 non-null float64
LDA_02                           39644 non-null float64
LDA_03                           39644 non-null float64
LDA_04                           39644 non-null float64
global_subjectivity              39644 non-null float64
global_sentiment_polarity        39644 non-null float64
global_rate_positive_words       39644 non-null float64
global_rate_negative_words       39644 non-null float64
rate_positive_words              39644 non-null float64
rate_negative_words              39644 non-null float64
avg_positive_polarity            39644 non-null float64
min_positive_polarity            39644 non-null float64
max_positive_polarity            39644 non-null float64
avg_negative_polarity            39644 non-null float64
min_negative_polarity            39644 non-null float64
max_negative_polarity            39644 non-null float64
title_subjectivity               39644 non-null float64
title_sentiment_polarity         39644 non-null float64
abs_title_subjectivity           39644 non-null float64
abs_title_sentiment_polarity     39644 non-null float64
shares                           39644 non-null int64
dtypes: float64(34), int64(26), object(1)
memory usage: 18.5+ MB

Decision Tree

y3 = target3
X3 = news.drop(['url', 'shares'], axis=1)

from sklearn.model_selection import train_test_split
X3_train, X3_test, y3_train, y3_test = train_test_split(X3, y3, test_size = 0.30, random_state = 2345)

from sklearn import tree 
from sklearn.tree import DecisionTreeClassifier 

model3 = tree.DecisionTreeClassifier() 
model3 = model3.fit(X3_train, y3_train)

model3

DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=None,
                       max_features=None, max_leaf_nodes=None,
                       min_impurity_decrease=0.0, min_impurity_split=None,
                       min_samples_leaf=1, min_samples_split=2,
                       min_weight_fraction_leaf=0.0, presort=False,
                       random_state=None, splitter='best')

from graphviz import Source
from IPython.display import display
from IPython.display import SVG

graph3 = Source(tree.export_graphviz(model3, out_file=None, feature_names=X3.columns, class_names=['default', 'no default'], filled = True))

display(SVG(graph3.pipe(format = 'svg')))

from sklearn.metrics import confusion_matrix 
from sklearn.metrics import accuracy_score

y_predict3 = model3.predict(X3_test) 

print(confusion_matrix(y3_test, y_predict3)) 
print(accuracy_score(y3_test, y_predict3)*100)

[[3043 2493]
 [2540 3818]]
57.684546830334625

YES!! I got 58% back by dropping "shares" as well!! Thank you Prof.Sadeghian! The accuracy score of Decision Tree here now says 100%, which means something is wrong. When I ran this code before, I got 58.11%.

Random Forest

from sklearn.ensemble import RandomForestClassifier 

clf3 = RandomForestClassifier()
clf3.fit(X3_train, y3_train)

/Users/hit028/opt/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
  "10 in version 0.20 to 100 in 0.22.", FutureWarning)

RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',
                       max_depth=None, max_features='auto', max_leaf_nodes=None,
                       min_impurity_decrease=0.0, min_impurity_split=None,
                       min_samples_leaf=1, min_samples_split=2,
                       min_weight_fraction_leaf=0.0, n_estimators=10,
                       n_jobs=None, oob_score=False, random_state=None,
                       verbose=0, warm_start=False)

y_predict3 = clf3.predict(X3_test) 

print(confusion_matrix(y3_test, y_predict3))
print(accuracy_score(y3_test, y_predict3)*100)

[[3598 1938]
 [2534 3824]]
62.4012106944678

By dropping "shares" brought the correct accuracy score here too! It's 62%

This time I got 99.9% accuracy here but I've got the accuracy score of Random Forest as 62% before.

import pandas as pd 
feature_importances3 = pd.DataFrame(clf3.feature_importances_, index = X3_train.columns,
columns = ['importance']).sort_values('importance', ascending = False)

feature_importances3

	importance
kw_avg_avg	0.044961
kw_max_avg	0.039708
timedelta	0.032124
LDA_02	0.031305
LDA_04	0.029921
kw_avg_min	0.029107
self_reference_min_shares	0.028792
LDA_01	0.028319
kw_avg_max	0.028124
LDA_00	0.028004
n_non_stop_unique_tokens	0.027467
global_rate_positive_words	0.027210
global_subjectivity	0.026168
self_reference_avg_sharess	0.025934
n_tokens_content	0.025914
LDA_03	0.025783
kw_max_min	0.025687
self_reference_max_shares	0.025296
n_unique_tokens	0.025202
average_token_length	0.024762
avg_positive_polarity	0.024755
global_sentiment_polarity	0.024405
kw_min_avg	0.024012
avg_negative_polarity	0.022835
global_rate_negative_words	0.022316
rate_negative_words	0.020523
num_hrefs	0.019427
rate_positive_words	0.019296
kw_min_max	0.016193
n_tokens_title	0.015332
title_sentiment_polarity	0.013984
max_negative_polarity	0.013780
num_imgs	0.013563
min_negative_polarity	0.013465
num_self_hrefs	0.013236
title_subjectivity	0.012743
min_positive_polarity	0.012274
abs_title_sentiment_polarity	0.011636
abs_title_subjectivity	0.011400
max_positive_polarity	0.010657
data_channel_is_entertainment	0.010132
num_keywords	0.009906
is_weekend	0.009642
num_videos	0.008023
kw_max_max	0.006387
kw_min_min	0.004276
data_channel_is_world	0.004203
data_channel_is_tech	0.003904
weekday_is_wednesday	0.003550
weekday_is_saturday	0.003390
weekday_is_thursday	0.003358
weekday_is_tuesday	0.003170
weekday_is_friday	0.002933
weekday_is_sunday	0.002546
weekday_is_monday	0.002519
data_channel_is_socmed	0.002361
data_channel_is_bus	0.002045
data_channel_is_lifestyle	0.001881
n_non_stop_words	0.000153

From above, the 3 most important features here are "shares", "kw_max_avg:Avg.keyword(max.shares)", and "LDA_02:Closeness to LDA topic 2". But when I ran this code earlier, it gave me "kw_avg_avg:Avg.keyword(avg.shares)" as the most importance feature instead of "shares".